Homeless youths, ages 18-23 years, are extremely vulnerable to poor health outcomes related particularly to substance use and sexual behaviors; adverse health outcomes include substance abuse and HIV/AIDs, often in combination. Although these youths' lives are characteristically unpredictable and the future looks bleak to them and to the larger society, they also have inherent strengths such as optimism, hope, self-efficacy, resilience, and gratitude that represent some of their psychological capital. Psychological capital has been shown to be a strong motivator of behavior. To date very few theory-driven strengths-based interventions have been developed and tested in this population in settings where they drop-in to seek health and human services. The specific aims of this application are to (1) determine the effectiveness of a self-reflective preventive intervention to promote responsible substance use and safe sex behaviors (prevent HIV/AIDS) in homeless youths, ages18-23 years; (2) determine whether psychological capital factors mediate the relationship between background risk factors (sexual orientation, duration of homelessness, sex abuse history, history of substance use, past sexual risk behaviors, and service utilization) and improved proximal outcomes (self-efficacy and intentions) and distal outcomes (responsible substance use, safer sex, life satisfaction); (3) determine the effect of pre-testing on interventin proximal outcomes (immediately following intervention) and distal outcomes (at an interval of 3 and 6 months following intervention); and an exploratory aim (4) to explore general and specific differences in proximal and distal outcomes of the intervention between participants in two geographically different sites (Austin, TX and Columbus, OH). We will recruit a total of 600 homeless youths from 2 drop-in centers in 2 geographic areas of the central US: Austin, TX and Columbus, OH. Using a Solomon four-group design with repeated measures, we will conduct a randomized controlled trial of a brief possible selves intervention (PSI) consisting of 6 brief individualized sessions that focus on generating images of one's possible self as feared or as best. Intervention participants will also receive weekly electronic messages to encourage them to move toward a goal of becoming their best possible self. Quantitative data will be collected before, immediately after the intervention, and at intervals of 3 and 6 months following intervention to test hypotheses about the effectiveness of the intervention on health behaviors and life satisfaction as well as the effects of pre-testing on these outcomes using structural equation modeling and ANCOVA statistics. The findings are expected to have a public health impact on further development of individualized services to promote healthy behaviors, healthy lifestyles and life satisfaction for vulnerable youths.